-
Notifications
You must be signed in to change notification settings - Fork 12.8k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Issue #5977 - Generalizing RawNullablePointer to RawForbiddenValue #31215
Conversation
r? @brson (rust_highfive has picked a reviewer for you, use r? to override) |
2a1317c
to
33b2674
Compare
assert_eq!(ix, 0); | ||
val | ||
RawForbiddenValue { payload_discr, .. } => { | ||
if _discr == payload_discr { |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
You can remove the _ here since the variable is now used.
\o/ seems good! Paves the way for more cool optimizations. |
I haven't managed to generalize |
@bors r+ |
📌 Commit 962221b has been approved by |
8ab79b1
to
6dec707
Compare
Fixed style and a test that was accidentally ripped from my next patch on the line. |
☔ The latest upstream changes (presumably #30448) made this pull request unmergeable. Please resolve the merge conflicts. |
Fixed. |
}, | ||
ty::TyRawPtr(..) | ty::TyInt(..) | ty::TyUint(..) => { | ||
path.push(0); | ||
Some(path) | ||
Some((path, C_null(type_of::sizing_type_of(cx, ty)))) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
ty
here would be the NonZero strict so I don't think that's what you want.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Not sure I understand your sentence, but I just realized that ty
was shadowed and that I should probably use the shadowed version instead. Is this the problem you were mentioning?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
s/strict/struct/ :P
So, the forbidden value will be compared against the discriminant field. Say we had Option<NonZero<int>>
, we're testing that that nested int is non-zero. Currently, it's returning C_null
of NonZero
(the passed in ty
to the function) as the forbidden value rather than of the int field (field_ty
).
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Got it, thanks. I obviously misunderstood that block.
r? @pnkfelix |
@luqmana I can put a |
…r> and Option<bool> r?luqmana
/me grumbles something about 100-chars limit. |
☔ The latest upstream changes (presumably #31417) made this pull request unmergeable. Please resolve the merge conflicts. |
@@ -438,23 +469,34 @@ struct Case<'tcx> { | |||
/// This represents the (GEP) indices to follow to get to the discriminant field | |||
pub type DiscrField = Vec<usize>; | |||
|
|||
fn find_discr_field_candidate<'tcx>(tcx: &ty::ctxt<'tcx>, | |||
fn find_discr_field_candidate<'a, 'tcx>(cx: &CrateContext<'a, 'tcx>, |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Please either add a doc comment explaining what the two parts of the returned tuple are and/or switch from using a tuple to using a struct with self-explanatory named fields
commits lgtm: |
@@ -1 +1 @@ | |||
Subproject commit 30f70baa6cc1ba3ddebb55b988fafbad0c0cc810 | |||
Subproject commit 91ff43c736de664f8d3cd351e148c09cdea6731e |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
hmm was this an accidental commit?
@Yoric I'd say this is definitely on the right track! Great stuff, sorry it took me so long to get around to looking seriously at it. If you rebase I suspect I'll just do a superficial review before r-plussing. (In other words, if another reviewer happens to get to it and wants to do r=pnkfelix, that will probably be fine.) |
Closing due to inactivity, but feel free to resubmit with a rebase! |
This is a WIP first step towards Issue #5977. Could someone (@luqman?) take a first look and tell me whether I'm heading in the right direction?